Back-drilling concentricity deviation compensation method and back-drilling concentricity deviation compensation system

Abstract

This application discloses a method and system for compensating for concentricity deviation in back-drilling, comprising: acquiring circuit board inspection data; screening and determining the accuracy of through holes based on the inspection data to generate a verification object set; when the verification object set is empty, directly generating target data for back-drilling processing and sending it to the circuit board processing equipment; when the verification object set is not empty, pushing the verification object set and its corresponding inspection data to a human-machine interface, displaying the through holes to be verified and their corresponding hole-level inspection images, and receiving manual operation until all through holes in the verification object set have been processed and passed the final deviation verification, generating the target data for back-drilling processing and sending it to the circuit board processing equipment for processing. This application improves the concentricity of back-drilled holes and through holes by generating target data for back-drilling processing through a closed-loop processing that combines automatic compensation for through hole position deviation with manual verification.

Description

Technical Field

[0001] This application belongs to the field of circuit board manufacturing methods or processes, specifically relating to a back-drill concentricity deviation compensation method and a back-drill concentricity deviation compensation system. Background Technology

[0002] With the development of applications such as artificial intelligence and high-performance computing, the reliability requirements for interlayer interconnects in products like AIPCBs have significantly increased. Back-drilling is used to remove excess stubs from vias, and its processing quality is highly dependent on the concentricity of the secondary drilled hole and the through-hole (or target hole center). Insufficient concentricity can lead to stub residue exceeding tolerances, damage to functional copper layers, and degradation of impedance and signal integrity, resulting in batch rework or even scrap. How to improve the concentricity of back-drilled holes and through-holes is a problem that urgently needs to be solved by those skilled in the art. Summary of the Invention

[0003] To address the aforementioned technical problems, this application provides a back-drilling concentricity deviation compensation method, comprising: acquiring circuit board inspection data, including via location information; based on the inspection data, screening and accuracy determination of vias to generate a verification object set; determining whether the verification object set is empty; when the verification object set is empty, generating target data for back-drilling based on the inspection data and sending the target data to the circuit board processing equipment for processing; when the verification object set is not empty, pushing the verification object set and its corresponding inspection data to a human-machine interface, displaying the vias to be verified and their corresponding hole-level inspection images, and receiving manual operation until all vias in the verification object set have been processed and passed the final deviation verification, generating target data for back-drilling and sending it to the circuit board processing equipment for processing.

[0004] In some embodiments of this application, generating a verification object set includes: dividing the through holes into a first type of hole and a second type of hole, and adding the second type of hole to the verification object set. The first type of hole includes: through holes that are successfully identified, have neighborhood reference conditions, and whose concentricity deviation is within a preset accuracy threshold range; the second type of hole includes: holes that failed to be identified, holes without neighborhood reference, and holes whose concentricity deviation does not meet the standard.

[0005] In some embodiments of this application, vias that fail to be identified during the detection process are marked as identification failures and added to the verification object set; a neighborhood range is defined for the identified vias, and neighboring vias within the neighborhood range are retrieved for each via; when no neighboring vias are retrieved or the number of neighboring vias is less than a preset value, the via is determined to be a via without a neighborhood reference and added to the verification object set; when the number of neighboring vias retrieved is greater than or equal to a preset value, automatic neighborhood compensation is performed on the via.

[0006] In some embodiments of this application, automatic neighborhood compensation includes: statistically analyzing the concentricity deviation values ​​of all adjacent vias within the neighborhood, and calculating the absolute difference between the components of the current via's concentricity deviation value and the neighborhood concentricity deviation reference value in each preset direction; when the absolute difference in any preset direction is greater than a preset difference threshold, correcting the concentricity deviation component of the current via in the preset direction to the corresponding component of the neighborhood concentricity deviation reference value; and when the absolute difference in all preset directions is not greater than the preset difference threshold, keeping the original concentricity deviation value of the current via unchanged.

[0007] In some embodiments of this application, after completing the automatic neighborhood compensation, the through holes are automatically judged for accuracy, including: calculating the concentricity deviation between the actual position and the theoretical position of each through hole; marking through holes with a concentricity deviation greater than a preset accuracy threshold as holes with unacceptable concentricity deviation and adding them to the verification object set; and classifying through holes with a concentricity deviation not greater than the preset accuracy threshold as first-class holes.

[0008] In some embodiments of this application, after pushing and displaying the set of objects to be reviewed on the human-computer interaction interface, the method further includes: loading the set of objects to be reviewed and the corresponding detection data, and displaying the through holes to be reviewed in a list in sequence; in response to the selection of a through hole to be reviewed in the list, loading the hole-level detection image corresponding to the hole, wherein the hole-level detection image is a locally magnified image of the area where the through hole is located, acquired by the vision inspection device, and superimposing a standard reference hole position mark (fixed) and a current hole position mark to be corrected (draggable and adjustable) on the hole-level detection image; comparing the standard reference hole position with the current hole position to be corrected, and determining the type of manual review operation for the current hole position to be corrected, wherein the operation type includes confirmation, skipping, or correction.

[0009] In some embodiments of this application, when the manual review operation type is correction, in response to the manual interactive adjustment operation of the current hole position to be corrected, the pixel deviation component between the current hole position to be corrected and the standard reference hole position in the preset direction is obtained; based on the pre-calibrated correspondence between pixel size and physical size, the pixel deviation component is converted into a physical offset compensation amount, and the physical offset compensation amount is used as the concentricity deviation compensation value of the through hole.

[0010] In some embodiments of this application, after determining the concentricity deviation compensation value, the physical offset compensation amount is updated and written into the detection data of the through-hole to update its concentricity deviation value, and the through-hole is marked as processed in the review object set; the remaining state of the review object set is scanned to determine whether there are any unprocessed through-holes in the review object set; if there are unprocessed through-holes, the step of displaying the through-holes to be reviewed in a list is returned; if there are no unprocessed through-holes, the step of sending target data is executed. Manual interactive adjustment operations include drag-and-drop movement or point fine-tuning operations; when a skip instruction is received, the current concentricity deviation value of the through-hole is kept unchanged, and the through-hole is directly marked as processed.

[0011] In some embodiments of this application, before sending the target data, a final deviation check is performed on all through holes marked as processed in the review object set: when a through hole marked as processed is checked and its current concentricity deviation value is still greater than a preset accuracy threshold, the marking status of the through hole in the review object set is reset to unprocessed, and the review information corresponding to the through hole is pushed back to the human-machine interface, and the manual correction step is re-executed; until all through holes in the review object set are marked as processed and the final check passes, the target data for back drilling is generated and sent.

[0012] This application also provides a back-drill concentricity deviation compensation system, comprising: a circuit board processing equipment for performing through-hole processing and back-drilling; a vision inspection device for inspecting through-holes and outputting inspection data; a host computer, wherein the vision inspection device is electrically connected to the host computer; the host computer is electrically connected to the circuit board processing equipment; the host computer acquires the inspection data, generates target data for back-drilling, and controls the circuit board processing equipment to apply the target data to process the back-drilled holes.

[0013] The back-drill concentricity deviation compensation method and system provided in this application achieve the following technical effects through a closed-loop processing combining automatic compensation for through-hole position deviation with manual verification: By combining automatic neighborhood compensation with manual verification, the accuracy and reliability of concentricity deviation compensation are significantly improved; holes without neighborhood references, failed identification holes, and holes with unacceptable concentricity deviations are uniformly included in the verification object set for manual confirmation or correction, effectively eliminating blind spots in automatic correction and ensuring the integrity of processing data; the detection table is associated with the hole-level image, and the pixel deviation is converted into physical offset compensation based on calibration parameters and written back to the detection data, providing accurate quantitative basis for manual correction and improving the traceability of correction results; the target data for back-drill processing is directly generated using the verified detection data as the authoritative data source, eliminating the need to re-execute the complete visual table processing, significantly shortening the closed-loop cycle from verification to processing, and ensuring that the manual correction results can be effectively converted into processing parameters; it supports multi-axis detection data management, separate verification, and unified generation of target data, making it suitable for multi-axis processing scenarios and improving data management efficiency. Finally, theoretical coordinate data of the back drill hole is generated based on the compensated through hole position, thereby effectively improving the concentricity of the back drill hole and the through hole. Attached Figure Description

[0014] To more clearly illustrate the technical solutions in the embodiments of this application, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0015] Figure 1 This is a schematic diagram of the overall flow of a back-drill concentricity deviation compensation method provided in an embodiment of this application;

[0016] Figure 2 This is a partial structural diagram of a circuit board processing equipment provided in an embodiment of this application;

[0017] Figure 2 The one-to-one correspondence between the component names and the attached labels is as follows: 10, spindle assembly; 20, worktable; 30, crossbeam; 40, base; 11, vision inspection device; 21, circuit board. Detailed Implementation

[0018] Various exemplary embodiments of this disclosure will now be described in detail with reference to the accompanying drawings. It should be noted that, unless specifically stated otherwise, the relative arrangement, numerical expressions, and values ​​of the components and steps set forth in these embodiments do not limit the scope of this disclosure. The following description of at least one exemplary embodiment is merely illustrative and is in no way intended to limit this application or its application or use. Techniques, methods, and apparatus known to those skilled in the art may not be discussed in detail, but where appropriate, such techniques, methods, and apparatus should be considered part of the specification. It should be noted that similar reference numerals and letters in the following drawings denote similar items; therefore, once an item is defined in one drawing, it need not be further discussed in subsequent drawings.

[0019] The specific embodiments of this disclosure are described below with reference to the accompanying drawings. In this document, terms such as "upper," "lower," "front," "back," "left," and "right" are used only to indicate the relative positional relationship between related parts, and not to limit the absolute position of these related parts. In this document, terms such as "first," "second," and "third" are used only to distinguish them from each other, and not to indicate degree of importance, order, or prerequisite for each other's existence. In this document, terms such as "equal," "same," "synchronous," "concentric," "simultaneously," "in sequence," "consistent," "symmetrical," "replicated," "center," and "center" are not strict mathematical and / or geometric limitations, and also include errors that are understandable to those skilled in the art and permissible in manufacturing or use.

[0020] The circuit board processing equipment in this application, such as Figure 2 As shown, the system includes a base 40, a crossbeam 30, a spindle assembly 10, and a worktable 20. The worktable 20 is mounted on the base 40, and the crossbeam 30 is mounted above the worktable 20. At least one spindle assembly 10 is slidably mounted on the crossbeam 30, and each spindle assembly 10 moves along a first direction on the crossbeam 30. The worktable 20 is on the base 40 and moves along a second direction. The worktable 20 has at least one processing station, and a circuit board 21 is placed on each processing station. Each spindle assembly 10 corresponds one-to-one with each circuit board 21 on each processing station. The tool held at the bottom of each spindle assembly 10 moves along a third direction to process the corresponding circuit board 21. A vision inspection device 11 is mounted on the spindle assembly 10. The first direction, the second direction, and the third direction are mutually perpendicular. In the embodiments of this application, the depth direction of the through hole and the back drill hole is the third direction.

[0021] In the embodiments of this application, the circuit board processing equipment can be implemented as a drilling equipment, a forming equipment, a milling machine, a drilling and milling integrated machine, etc., and there is no limitation herein. In the embodiments of this application, at least one spindle assembly is slidably connected to the crossbeam of the circuit board processing equipment. The number of spindle assemblies can be one, two, three, four, five, six, ten, twelve, etc., and the structure and function of each spindle assembly are basically the same, and there is no limitation herein.

[0022] This application provides a method for compensating for concentricity deviation in back drilling, comprising: acquiring detection data of a circuit board, the detection data including the position information of through holes; based on the detection data, screening and accuracy determination of through holes to generate a set of verification objects; determining whether the set of verification objects is empty; when the set of verification objects is empty, generating target data for back drilling based on the detection data, and sending the target data to the circuit board processing equipment for processing; when the set of verification objects is not empty, pushing the set of verification objects and its corresponding detection data to a human-machine interface, displaying the through holes to be verified and their corresponding hole-level detection images, and receiving manual operation until all through holes in the set of verification objects have been processed and passed the final deviation verification, generating target data for back drilling, and sending it to the circuit board processing equipment for processing.

[0023] Specifically, after scanning and inspecting the through-holes on the circuit board, the vision inspection device generates structured inspection data. This data includes the location information of the through-holes detected by the vision inspection device, such as hole coordinates, hole diameter, and concentricity deviation. For multi-axis machining scenarios, the inspection data is managed separately by axis number, with each axis corresponding to independent inspection data and image paths. After acquiring the inspection data, the host computer performs filtering and accuracy judgment, including holes requiring manual intervention into the review object set. When the review object set is empty, it indicates that the compensation amount of all through-holes has met the accuracy requirements in the automatic process, and the host computer directly generates target data using the current inspection data as the authoritative data source. When the review object set is not empty, the manual review interaction process begins. During the manual review, the host computer pushes the review object set and its corresponding inspection data to the human-machine interface, displays the through-holes to be reviewed and their corresponding hole-level inspection images, and receives manual operations. The operator visually judges the deviations and confirms or corrects them. After the manual operation is completed, the updated inspection data still serves as the authoritative data source to directly generate target data, without re-executing the complete inspection and recognition process. Finally, the target data is sent to the circuit board processing equipment, which adjusts the drilling positions based on the compensation information to process the back drill holes. This combination of automatic correction and manual verification ensures that only verified compensation data enters the processing stage, thereby improving the accuracy of back drill hole concentricity control and process controllability.

[0024] In some embodiments of this application, generating a review object set includes: dividing the through holes into a first type of hole and a second type of hole, and adding the second type of hole to the review object set.

[0025] Specifically, the host computer categorizes all through-holes into two types based on preset rules. The first type refers to through-holes that can directly enter the subsequent target data generation process, such as those that have been successfully identified, have neighborhood reference conditions, and whose concentricity deviation is within a preset accuracy threshold. For the first type of through-hole, their concentricity deviation already meets the accuracy requirements, and they can directly enter the subsequent processing data generation process. The second type refers to through-holes that are unsuitable or cannot be reliably processed by the automatic compensation algorithm, such as through-holes that failed to be identified, those without neighborhood references, and those whose concentricity deviation does not meet the standard. For the second type of through-hole, the host computer adds them to the review object set, awaiting manual review. By dividing through-holes into two categories and processing them separately, it is possible to distinguish normal through-holes to improve efficiency, while also allowing for manual review of risky through-holes to ensure quality, achieving a balance between efficiency and accuracy.

[0026] In some embodiments of this application, the first type of hole includes: a through hole that is successfully identified, has a neighborhood reference condition, and has a concentricity deviation within a preset accuracy threshold range; the second type of hole includes: a hole that fails to be identified, a hole without a neighborhood reference, and a hole whose concentricity deviation does not meet the standard.

[0027] The first type of hole must simultaneously meet three conditions: First, successful identification, meaning the detection equipment can successfully identify the position and concentricity deviation value of the through hole; second, having a neighborhood reference condition, meaning the through hole has a preset number of neighboring through holes within a preset neighborhood range, providing reference information for automatic compensation; and third, the concentricity deviation is within a preset accuracy threshold range, meaning that after automatic compensation correction, the concentricity deviation value of the through hole meets the process accuracy requirements. Only through holes that simultaneously meet the above three conditions are classified as first-type holes and directly enter the subsequent processing data generation process.

[0028] The second category of holes includes three types: failed-to-identify holes (those that the detection equipment failed to identify due to missing or unreliable concentricity deviation values); holes without neighboring references (those that, although successfully identified, have no adjacent holes within a preset neighborhood or whose number of adjacent holes is less than a preset value, making automatic neighborhood compensation impossible); and holes with concentricity deviation exceeding the standard (those whose concentricity deviation value still exceeds the preset accuracy threshold after automatic neighborhood compensation). All three types of holes are added to the review object set for manual confirmation or correction by the operator. By clearly defining the specific criteria for classifying the two types of holes, the first type of holes are used directly, while risky holes are uniformly included in manual review, reducing the possibility of missed or incorrect identifications.

[0029] In some embodiments of this application, through holes that are not identified during the detection process are marked as identification failure holes and added to the verification object set.

[0030] Specifically, during the scanning and inspection process, the visual inspection device may fail to identify certain through holes due to poor lighting conditions, hole wall contamination, or abnormal hole shape. In the inspection data, the identification status of the corresponding hole is marked as "unidentified," and the concentricity deviation value is missing. The host computer traverses the inspection data, marks the through holes with the identification status of "unidentified" as identification failure holes, and directly adds them to the review object set for manual secondary review.

[0031] Furthermore, for identified failed holes, the host computer can calculate a reference concentricity deviation value based on the transformation relationship of the positioning marks and write it into the detection data as an initial reference value for manual review. For example, by using the positioning hole to calculate the transformation matrix from theoretical design coordinates to actual image coordinates, the theoretical position of the hole is mapped to image coordinates and compared with the target hole center position to obtain the reference deviation value. Although this reference value may not be precise, it can provide the operator with an initial judgment basis. In this way, identified failed holes will not be missed due to data loss, but will be forcibly included in the manual review process, ensuring the integrity of the processing data. At the same time, the reference deviation value provides an initial reference for manual review, improving review efficiency.

[0032] In some embodiments of this application, a neighborhood range is defined for the identified vias, and neighboring vias within the neighborhood range are retrieved for each via. When no neighboring vias are retrieved or the number of neighboring vias is less than a preset value, the via is determined to be a via without a neighborhood reference and is added to the verification object set. When the number of neighboring vias retrieved is greater than or equal to the preset value, automatic neighborhood compensation is performed on the via.

[0033] Specifically, for each identified via, the host computer defines a preset neighborhood range centered on it. This neighborhood range can be a square or circular area centered on the current via, or it can be a dynamically determined area based on the via spacing. Other vias are searched within this neighborhood. If no neighboring vias are found, or the number of neighboring vias is less than a preset value, it indicates that the via is in an isolated position with no available neighborhood information. In this case, it is classified as a via without a neighborhood reference and added to the verification object set without automatic compensation. If the number of neighboring vias found is greater than or equal to the preset value, automatic neighborhood compensation is performed on the via. Through this neighborhood search and via without a neighborhood reference determination, automatic compensation is only performed when a neighborhood reference exists, avoiding errors that might be introduced by blindly correcting without a reference.

[0034] In some embodiments of this application, automatic neighborhood compensation includes: statistically analyzing the concentricity deviation values ​​of all adjacent vias within the neighborhood, and calculating the absolute difference between the components of the current via's concentricity deviation value and the neighborhood concentricity deviation reference value in each preset direction; when the absolute difference in any preset direction is greater than a preset difference threshold, correcting the concentricity deviation component of the current via in the preset direction to the corresponding component of the neighborhood concentricity deviation reference value; and when the absolute difference in all preset directions is not greater than the preset difference threshold, keeping the original concentricity deviation value of the current via unchanged.

[0035] Specifically, when an identified via exists within a neighborhood of a preset number of neighboring vias, the host computer statistically analyzes the concentricity deviation values ​​of all neighboring vias and calculates a neighborhood concentricity deviation reference value. The reference value can be the average, median, or other statistical value of the concentricity deviation values ​​of neighboring vias. The deviation values ​​in the X and Y directions are statistically analyzed to obtain the neighborhood reference value. Then, the absolute difference between the current via's concentricity deviation value and the neighborhood concentricity deviation reference value in each preset direction is calculated, and each absolute difference is compared with a preset difference threshold. When the absolute difference in any preset direction is greater than the preset difference threshold, the concentricity deviation component of the current via in that preset direction is corrected to the corresponding component of the neighborhood concentricity deviation reference value; when the absolute differences in all preset directions are not greater than the preset difference threshold, the original value remains unchanged. This neighborhood statistical consistency correction method can automatically eliminate sporadic errors and local anomalies in the detection data, making the overall deviation data smoother and more reasonable, and improving the accuracy and reliability of automatic compensation.

[0036] In some embodiments of this application, after completing automatic neighborhood compensation, automatic accuracy determination is performed on the through holes for which automatic neighborhood compensation has been completed, including: calculating the concentricity deviation between the actual position and the theoretical position of each through hole; marking through holes with a concentricity deviation greater than a preset accuracy threshold as holes with unacceptable concentricity deviation and adding them to the verification object set; and classifying through holes with a concentricity deviation not greater than the preset accuracy threshold as first-class holes.

[0037] Specifically, after automatic neighborhood compensation is completed, the host computer performs accuracy determination on each through-hole that has completed automatic neighborhood compensation. First, it calculates the concentricity deviation between the actual position and the theoretical position of each through-hole. The deviation can be a corrected deviation or the original deviation (for uncorrected holes), typically including components in the X and Y directions. Then, it compares the concentricity deviation with a preset accuracy threshold. If the deviation is greater than the preset accuracy threshold, the through-hole is marked as a hole with unacceptable concentricity deviation and added to the verification set; if the deviation is not greater than the preset accuracy threshold, the through-hole meets the accuracy requirements and is classified as a first-class hole, which can directly enter the subsequent processing flow. Through automatic accuracy determination, through-holes that still do not meet the accuracy requirements after automatic compensation can be automatically screened out and included in manual verification, thereby ensuring that the compensation data used for final processing all meet the process accuracy requirements.

[0038] In some embodiments of this application, after pushing and displaying the set of objects to be reviewed on the human-computer interaction interface, the method further includes: loading the set of objects to be reviewed and the corresponding detection data, and displaying the through holes to be reviewed in a list in sequence; in response to the selection of a through hole to be reviewed in the list, loading the hole-level detection image corresponding to the hole, wherein the hole-level detection image is a locally magnified image of the area where the through hole is located, acquired by the vision inspection device, and superimposing the standard reference hole position and the current hole position to be corrected on the hole-level detection image; comparing the standard reference hole position and the current hole position to be corrected, and determining the type of manual review operation for the current hole position to be corrected, wherein the operation type includes confirmation, skipping, or correction.

[0039] Specifically, when the set of objects to be reviewed is not empty, the host computer loads the set of objects to be reviewed and the corresponding test data. All through-holes to be reviewed are displayed in a list format on the human-machine interface. List items can include information such as hole identifier, current concentricity deviation value, and marker type. The operator clicks on a hole in the list, and the host computer loads the corresponding hole-level test image. The hole-level test image is a magnified local image of the through-hole acquired in the testing equipment. The host computer overlays a standard reference hole position (fixed) and the current hole position to be corrected (adjustable by dragging) onto the image. By comparing the relative positions of the two markers, the operator can intuitively determine the direction and magnitude of the concentricity deviation of the current through-hole, and thus determine the type of manual review operation for the current hole position to be corrected. The operation type includes confirmation, skipping, or correction. Through the linked display of the list and image, the test data and hole-level images are associated one-to-one. The operator can quickly locate risky holes and intuitively judge the deviation on the image, thereby improving the efficiency and accuracy of manual review and solving the problem of the separation between test data and hole-level images.

[0040] In some embodiments of this application, when the manual review operation type is correction, the step of determining the concentricity deviation compensation value includes: in response to a manual interactive adjustment operation on the current hole position to be corrected on the hole-level inspection image, obtaining the pixel deviation component in a preset direction between the current hole position to be corrected and the standard reference hole position after adjustment; based on the pre-calibrated correspondence between pixel size and physical size (including camera micro angle, resolution and coordinate quadrant information), converting the pixel deviation component into a physical offset compensation amount, and using the physical offset compensation amount as the concentricity deviation compensation value of the through hole.

[0041] Specifically, when the operator determines the manual review operation type as correction, manual interactive adjustments can be made to the current hole position to be corrected on the hole-level inspection image. For example, the hole position can be adjusted to coincide with the desired target hole center position by dragging with the mouse or fine-tuning with the keyboard. After adjustment, the host computer obtains the pixel deviation components (including pixel differences in the X and Y directions) between the adjusted hole position and the standard reference hole position. Then, based on pre-calibrated parameters, the host computer converts the pixel deviation components into a physical offset compensation amount. The pre-calibrated parameters may include camera micro-angle, resolution, and coordinate quadrant information. The specific conversion process is as follows: first, the pixel deviation component is rotated and compensated according to the camera micro-angle; then, it is multiplied by the resolution to obtain the physical offset; finally, the direction sign is adjusted according to the coordinate quadrant convention to obtain the physical offset compensation amount. The physical offset compensation amount is the concentricity deviation compensation value of the through hole. This method establishes a traceable conversion link from image pixel difference to physical compensation amount, providing accurate quantitative basis for manual correction operations, reducing the subjectivity and arbitrariness of manual correction, and improving the accuracy and repeatability of the correction results.

[0042] In some embodiments of this application, after determining the concentricity deviation compensation value, the method further includes: updating the physical offset compensation amount in the detection data of the via to update its concentricity deviation value, and marking the via as processed in the review object set; scanning the remaining state of the review object set to determine whether there are any unprocessed vias in the review object set; when there are unprocessed vias, returning and executing the step of sequentially displaying the vias to be reviewed in a list form; when there are no unprocessed vias, executing the step of sending target data.

[0043] Specifically, after manual correction is completed, the host computer writes the calculated physical offset compensation amount into the through-hole's detection data, updating its concentricity deviation value. Simultaneously, the through-hole is marked as processed in the review object set. Subsequently, the host computer scans the remaining status of the review object set to check for any unprocessed through-holes. If any exist, the process returns to the list display step, showing the next hole to be reviewed, and the operator continues processing; if none exist, the process proceeds to the next step, generating target data based on the updated detection data and sending it to the processing equipment. By iterating through all holes in the review object set, it ensures that each hole to be reviewed is processed, avoiding omissions; simultaneously, marking processed status facilitates tracking of review progress and improves process controllability.

[0044] In some embodiments of this application, the manual interactive adjustment operation includes: in response to a drag-and-move or point fine-tuning operation performed on the hole position to be corrected on the hole-level detection image, to determine the final pixel position of the hole position to be corrected; when a skip instruction for the current through hole is received in the human-computer interaction interface, the operation of obtaining pixel deviation components and calculating physical offset compensation amount is not performed on the through hole, the current concentricity deviation value of the through hole is kept unchanged, and the through hole is directly marked as processed.

[0045] Specifically, manual adjustment operations can include two methods: drag-and-drop and fine-tuning. Drag-and-drop refers to the operator dragging the marker of the hole to be corrected using the mouse to move it to the target position; fine-tuning refers to the operator making small adjustments to the hole using the keyboard arrow keys or by inputting a fine-tuning value. Both methods can determine the final pixel position of the hole to be corrected. Furthermore, when the operator determines that the hole does not need correction or needs to be left unprocessed, they can issue a skip command in the human-computer interaction interface. Upon receiving the skip command, the host computer does not perform pixel deviation component acquisition or physical offset compensation calculation operations on the through-hole, retains the current concentricity deviation value of the through-hole unchanged, and directly marks the through-hole as processed. This allows the operator to quickly skip holes that do not require correction and concentrate on processing the holes that truly need correction, thereby improving the efficiency of manual review and the user experience.

[0046] In some embodiments of this application, before sending the target data, a final deviation check is performed on all through holes marked as processed in the review object set: when a through hole marked as processed is checked and its current concentricity deviation value is still greater than a preset accuracy threshold, the marking status of the through hole in the review object set is reset to unprocessed, and the review information corresponding to the through hole is pushed back to the human-machine interface, and the manual correction step is re-executed; until all through holes in the review object set are marked as processed and the final check passes, the target data for back drilling is generated and sent.

[0047] Specifically, after all through holes in the verification object set are marked as processed, the host computer performs a final deviation check on the updated inspection data before generating the target data. Specifically, it iterates through the current concentricity deviation values ​​of all processed holes in the verification object set and compares them with a preset accuracy threshold. If the current concentricity deviation value of a hole is found to be greater than the preset accuracy threshold, it indicates that although the hole has undergone manual processing, it still does not meet the accuracy requirements. In this case, the host computer re-pushes the verification information corresponding to that through hole to the human-machine interface and re-executes the manual correction step. This process can be iterated or repeated until all through holes in the verification object set are marked as processed and the final verification passes, at which point the target data for back-drilling is generated and sent. This closed-loop mechanism of final deviation verification prevents holes that still do not meet the accuracy standards after manual correction from being missed, ensuring the quality of the target data and thus improving the first-pass success rate and product yield of back-drilling.

[0048] This application also provides a back-drill concentricity deviation compensation system, comprising: a circuit board processing equipment for performing through-hole processing and back-drilling; a vision inspection device for inspecting through-holes and outputting inspection data; a host computer, wherein the vision inspection device is electrically connected to the host computer; the host computer is electrically connected to the circuit board processing equipment; the host computer acquires the inspection data, generates target data for back-drilling, and controls the circuit board processing equipment to apply the target data to process the back-drilled holes.

[0049] Specifically, the circuit board processing equipment can adjust the drilling position according to the compensation amount in the target data sent by the host computer to process back-drilled holes. The vision inspection device can be a vision inspection system, which is set on the circuit board processing equipment. It can inspect through holes online in real time and output inspection data containing information such as hole position coordinates and concentricity deviation values. The host computer obtains the inspection data, calculates the deviation between the actual coordinates and the theoretical coordinates, determines the theoretical coordinate data of the back-drilled hole, and outputs the target data. The vision inspection device can also be an offline device, such as an optical inspection device or an X-ray inspection device, used to perform imaging inspection on the through holes on the circuit board, outputting inspection data containing information such as hole position coordinates and concentricity deviation values. The host computer obtains the inspection data, calculates the deviation between the actual coordinates and the theoretical coordinates, determines the theoretical coordinate data of the back-drilled hole, and outputs the target data. The host computer communicates with the circuit board processing equipment and the vision inspection device, and can be an industrial PC, industrial control computer, or server, etc. The host computer runs the software corresponding to the method described in this application, acquires the inspection data output by the vision inspection device, performs steps such as screening, accuracy judgment, manual review and interaction, and target data generation, and sends the target data to the circuit board processing equipment to control its processing of back-drilled holes. By integrating inspection, compensation calculation, manual review and processing control into a complete system, an automated closed loop for improving the concentricity of back-drilling is realized, which helps to improve processing accuracy and efficiency and is suitable for mass production of high-precision circuit boards such as AIPCB.

[0050] In some embodiments of this application, the host computer in the above system is used to execute the back-drill concentricity deviation compensation method of any of the foregoing embodiments. Specifically, the host computer stores a computer program, which, when executed by the host computer's processor, implements all steps of the back-drill concentricity deviation compensation method. By executing this method, the host computer can automatically complete processes such as acquiring and filtering detection data, generating a set of verification objects, engaging in manual verification interaction, final verification, and generating and distributing target data. By integrating the method of this application into the host computer, the entire system possesses automation and intelligence capabilities, lowers the technical threshold for operators, improves the consistency and reliability of back-drill concentricity compensation, and facilitates deployment and promotion in production lines.

[0051] In the embodiments of this application, through holes are primary holes, and back-drilled holes are secondary holes. Secondary holes and primary holes can be in the same direction or opposite directions; no limitation is made here. Through holes can be holes with depth or PADs including preset patterns. In the embodiments of this application, target data are program files loaded by circuit board processing equipment. The circuit board processing equipment loads different target data to process different types of circuit boards.

[0052] Example

[0053] This embodiment uses an AIPCB circuit board as the processing object. This board has multiple processing axes (e.g., Z1, Z2), such as... Figure 2 As shown, the vision inspection device 11 is a vision inspection system for circuit board processing equipment, which is a mechanical drilling equipment for processing through holes and back-drilled holes. The host computer runs the software corresponding to the method described in this application and communicates with the vision inspection device 11 and the circuit board processing equipment. All preset parameters (neighborhood size D, difference threshold Tnb, accuracy tolerance Tol, camera micro-angle θ, resolution res, coordinate quadrant FV, etc.) are configurable items and are preset according to actual process conditions. In this embodiment, Tol can be on the order of approximately 0.05 mm, but this does not constitute a limitation on the scope of protection of this application.

[0054] Combined with appendix Figure 1 To be continued Figure 2 The diagram illustrates in detail the process of applying the back-drill concentricity deviation compensation method in the back-drill concentricity deviation compensation system. In the following steps, SPOF represents the concentricity deviation of the through hole, i.e., the compensation amount; the subscripts x and y represent the components in the X and Y directions, respectively; SPOFx represents the compensation amount in the X direction, and SPOFy represents the compensation amount in the Y direction.

[0055] S100: Acquire detection data.

[0056] After completing the hole scanning inspection, the vision inspection device 11 generates structured inspection data. This file contains hole identifiers (hole number, axis number, block number), compensation amounts (SPOFx, SPOFy), recognition and judgment status, and optional remarks for multiple holes. For multi-axis machining scenarios, the inspection data is managed separately by axis number, with each axis corresponding to independent inspection data and image paths. For example, the Z1 axis corresponds to the inspection data Z1_Result.txt and the image directory Z1_Images / , and the Z2 axis corresponds to Z2_Result.txt and Z2_Images / . The host computer reads the inspection data and loads it into memory as the basis for subsequent screening and verification.

[0057] S120: Handling holes that fail to be identified.

[0058] The system iterates through all holes in the detection data. For holes with an identification status of "unidentified," a reference SPOF calculation is performed. Specifically, the host computer uses image processing algorithms to calculate the transformation matrix from theoretical design coordinates to actual image coordinates based on the hole location information. This transformation matrix, denoted as M, maps the theoretical position (design coordinates) of the failed-identified hole to image coordinates, obtaining the estimated position of the hole on the image. This estimated position is compared with the target hole center position (e.g., the through-hole detection position) to obtain the reference SPOF (SPOFx_ref, SPOFy_ref). This reference SPOF is written into the entry for the corresponding hole in the detection data, giving the hole an initial compensation value. This reference SPOF serves only as an initial reference for manual review. If a skip instruction is received during manual review, the system defaults to this reference value. Simultaneously, the failed-identified hole is added to the review object set R.

[0059] Wherein, the transformation matrix M can be an affine transformation matrix, and its general form is:

[0060] M = [abc; def; 0 0 1]

[0061] The transformation relationship between theoretical coordinates (x_theory, y_theory) and image coordinates (x_img, y_img) is as follows:

[0062] x_img = a·x_theory + b·y_theory + c

[0063] y_img = d·x_theory + e·y_theory + f

[0064] S140: Perform screening and accuracy determination on successfully identified holes.

[0065] This step includes the following sub-steps:

[0066] S141: Define the neighborhood boundaries.

[0067] For each successfully identified hole i, the host computer delineates a preset circular neighborhood range centered on that hole, with the neighborhood radius denoted as R (for example, R = D / 2, where D is a preset neighborhood diameter parameter that can be configured according to the hole spacing distribution).

[0068] S143: Search for neighboring holes.

[0069] Within the defined neighborhood, other identified holes are retrieved to form a neighboring hole set N(i), where i represents the number of the currently retrieved hole. Specifically, the retrieval method is as follows: traverse all other successfully identified holes in the detection data, calculate the distance to the current hole, and designate holes with a distance less than or equal to R as neighboring holes.

[0070] S145: Determine whether the number of adjacent holes is less than the preset value.

[0071] Determine if the number of neighboring holes N(i) is less than a preset value. If N(i) is less than the preset value, execute S147; if N(i) is greater than or equal to the preset value, execute S149.

[0072] S147: Determination and processing of no neighboring reference holes.

[0073] If the number of neighboring holes N(i) is less than a preset value, meaning there are no neighboring holes or the number of neighboring holes is insufficient within the preset neighborhood, then hole i is determined to be a hole without a neighborhood reference. For holes without a neighborhood reference, its original compensation amount remains unchanged, and the hole without a neighborhood reference is added to the verification object set R. This hole does not undergo automatic neighborhood compensation, and the process jumps to S180 to perform the verification object set empty processing.

[0074] S149: Statistical neighborhood SPOF statistic.

[0075] If the number of neighboring holes N(i) is greater than or equal to a preset value, then the neighborhood SPOF statistic S is calculated. The statistic S can be calculated using either the mean or the median, for the X and Y directions respectively, yielding S_x and S_y. When using the mean, the formula for calculating the neighborhood SPOF statistic is:

[0076] S_x = (1 / n) · Σ_{j∈N(i)} SPOFx_j

[0077] S_y = (1 / n) · Σ_{j∈N(i)} SPOFy_j

[0078] When using the median, the formula for calculating the neighborhood SPOF statistic is:

[0079] S_x = median({SPOFx_j | j∈N(i)})

[0080] S_y = median({SPOFy_j | j∈N(i)})

[0081] Where n represents the number of neighboring holes in the neighboring hole set N(i), j represents the number of the neighboring hole, and SPOFx_j and SPOFy_j represent the compensation amount of the j-th neighboring hole in the X and Y directions, respectively.

[0082] S151: Neighborhood Consistency Comparison and Correction.

[0083] Calculate the absolute difference between the current orifice compensation amount and the components of the statistical quantity S in each direction, and compare them with the preset difference threshold Tnb. The judgment criterion for difference comparison is:

[0084] |SPOFx_i − S_x| > Tnb

[0085] |SPOFy_i − S_y| > Tnb

[0086] When any of the above conditions are met, the compensation amount in the direction corresponding to the current hole is corrected to the neighborhood statistics, that is:

[0087] SPOFx_i ← S_x, when |SPOFx_i − S_x| > Tnb

[0088] SPOFy_i ← S_y, when |SPOFy_i − S_y| > Tnb

[0089] Otherwise, the original value remains unchanged. After the correction or hold operation is completed, the screening and automatic neighborhood compensation for that hole ends. This automatic neighborhood compensation process can eliminate sudden changes in compensation amount caused by detection noise or local interference, making the overall compensation data more consistent with the neighborhood consistency rule. Then, it proceeds to S160 for automatic accuracy determination.

[0090] S160: Automatic accuracy determination and verification object set generation.

[0091] After completing automatic neighborhood compensation, automatic accuracy determination is performed on the vias (excluding vias without neighborhood references) that have completed automatic neighborhood compensation. The compensation amount for each via is compared with the preset accuracy tolerance Tol, and the determination criteria are as follows:

[0092] |SPOFx| ≤ Tol and |SPOFy| ≤ Tol

[0093] If the SpOFx or SpOFy of a hole exceeds the range, it is marked as a hole with unacceptable concentricity deviation and added to the verification object set R; if both conditions are met, the hole is classified as a first-class hole and can directly enter the subsequent processing data generation process.

[0094] Combining the failed identification holes added in S120 and the non-neighbor reference holes added in S140, a final verification object set R is generated. The verification object set R must contain at least one of the following: failed identification holes, holes with unacceptable concentricity deviation, or holes without neighbor reference. The host computer determines whether the verification object set R is empty to decide the subsequent process flow.

[0095] S180: Determine if the set of objects to be reviewed is empty and split the flow.

[0096] The host computer checks if the set of objects to be reviewed, R, is empty. If R is empty, it means that all through holes have met the accuracy requirements in the automatic screening process and no manual intervention is required. At this time, it jumps directly to S220, generates target data based on the current detection data, and then executes S240 to issue processing. If R is not empty, it means that there are risk holes that require manual intervention, and enters S200 to perform manual review and interactive correction.

[0097] S200: Manual review and interactive correction.

[0098] When R is not empty, the following sub-steps are executed:

[0099] S201: Related Display.

[0100] When the verification object set R is not empty, the verification object set and its corresponding detection data are pushed to the human-machine interface, displaying the through holes to be verified and their corresponding hole-level detection images. Specifically, the host computer loads the verification object set R and its corresponding detection data, and associates the holes to be verified with their corresponding hole-level detection images according to axis number, block number, and hole number. For multi-axis scenarios, the images are displayed in pages according to axis number, and the hole list within each axis is displayed independently. The interface displays the detection table and hole-level images side by side synchronously, allowing the operator to view both the table data and the images simultaneously without switching interfaces. When the operator selects a hole to be verified in the list, the host computer loads the corresponding Blk / Hole cropped image of that hole in the image area. This image is a magnified image of the hole acquired by the vision inspection device 11. The host computer overlays a standard reference hole position mark (point 2, fixed) and a current hole position mark to be corrected (point 1, draggable) onto the image. Point 2 represents the theoretical hole center or target hole center (such as the through hole detection position), and point 1 represents the current through hole center position identified or calculated by the vision inspection. The relative positions of the two markers visually represent the direction and magnitude of the current concentricity deviation of the hole.

[0101] S203: Receive manual operation.

[0102] Based on the hole-level image and compensation information, the operator determines the type of manual review operation for the current hole to be corrected, and performs one of the following operations on the current hole:

[0103] Confirm: If the current hole compensation amount is deemed reasonable, click Confirm. If the compensation amount remains unchanged, the hole is marked as processed.

[0104] Skip: When a skip command is received for the current via in the human-machine interface, the pixel deviation component acquisition and physical offset compensation calculation are not performed on the via. The current concentricity deviation value of the via remains unchanged, and the via is directly marked as processed. When the operator determines that the current compensation reference value of the current via (especially the via that failed to be identified) is reasonable or does not need adjustment, the operator can click skip. At this time, the host computer retains the existing compensation value in the detection data of the via (for the via that failed to be identified, the reference SPOF calculated in S120 is retained; for other vias, the neighborhood automatic compensation or the original detection value is retained), and only marks it as processed.

[0105] Correction: When the manual review operation type is correction, the operator can drag the current hole position mark to be corrected on the image to coincide with the standard reference hole position mark by dragging the mouse or fine-tuning the keyboard.

[0106] S205: Pixel Deviation Conversion and Write-back.

[0107] When the manual review operation type is correction, the host computer obtains the pixel deviation components between the standard reference hole position mark (point 2) and the adjusted current hole position mark to be corrected (point 1) on the image, denoted as (Δu_px, Δv_px)... where Δu_px represents the pixel deviation in the horizontal direction of the image, and Δv_px represents the pixel deviation in the vertical direction of the image. The pixel deviation is converted into an SPOF increment based on pre-calibrated parameters. Calibration parameters include camera micro-angle θ, resolution res, and coordinate quadrant information. Camera micro-angle θ is used to correct small rotational deviations between the image coordinate system and the physical coordinate system; resolution res represents the physical size corresponding to each pixel; coordinate quadrant information is used to determine whether the coordinate axis directions need to be reversed. Camera parameters can come from the batch CamParam configuration file, such as those containing fields like PixelsPerMM and ThetaArc. The conversion process is as follows: First, the pixel deviation is rotated and compensated according to the camera micro-angle θ to obtain the rotationally corrected deviation components Δx′ and Δy′, calculated using the following formula:

[0108] Δx′ = Δu_px·cosθ − Δv_px·sinθ

[0109] Δy′ = Δu_px·sinθ + Δv_px·cosθ

[0110] Here, cosθ and sinθ represent the cosine and sine values ​​of the camera's micro-angle θ, respectively.

[0111] Multiply by the resolution res to get the physical offset. The calculation formula is:

[0112] ΔSPOFx = Δx′·res

[0113] ΔSPOFy = Δy′·res

[0114] Finally, adjust the direction sign according to the coordinate quadrant convention. When the image coordinate Y direction is opposite to the physical Y direction, ΔSPOFy is inverted. The coordinate quadrant convention can be configured as FV=1 or FV=4.

[0115] The obtained ΔSPOFx and ΔSPOFy are superimposed with the original compensation amount of the current hole to obtain the updated compensation amount. The calculation formula is as follows:

[0116] SPOFx_new = SPOFx_old + ΔSPOFx

[0117] SPOFy_new = SPOFy_old + ΔSPOFy

[0118] Wherein, SPOFx_old and SPOFy_old represent the compensation amounts in the X and Y directions of the current hole before correction, respectively, and SPOFx_new and SPOFy_new represent the compensation amounts after update.

[0119] Before writing back the test data, the host computer automatically generates a backup file (e.g., .bak) to record the data state before modification, facilitating traceability and documentation. Then, the updated compensation amount is written into the corresponding entry in the test data, and the hole is marked as processed. Simultaneously, the hole-level image, overlaid with the standard reference hole position mark and the adjusted current hole position mark to be corrected, is saved as a separate marked correction image file for quality traceability and manual review recording. Optionally, a secondary tolerance judgment can be performed on the updated compensation amount, i.e., checking again whether |SPOFx|≤Tol and |SPOFy|≤Tol, and the judgment result is saved.

[0120] S207: Traversal processing.

[0121] After processing each hole, the host computer scans the remaining status of the review object set R to determine if there are any unprocessed holes. If so, it returns to S201 to continue processing the next hole to be reviewed; otherwise, it proceeds to S209.

[0122] S209: Final deviation verification.

[0123] Before generating the target data, the host computer performs a final deviation check on the updated detection data. It iterates through the current compensation values ​​of all processed holes in the verification object set R and compares them with the preset accuracy tolerance Tol, using the following criteria:

[0124] |SPOFx| ≤ Tol and |SPOFy| ≤ Tol

[0125] If the compensation amount of a processed hole is found to still exceed the tolerance range (i.e., |SPOFx| > Tol or |SPOFy| > Tol), the marking status of that hole in the review object set is reset to "unprocessed," and the review information corresponding to that hole is pushed back to the human-computer interaction interface, and the manual correction steps S201 to S207 are re-executed. This process can be iterated until all through holes in the review object set are marked as "processed" and the final verification passes.

[0126] S220: Generate target data.

[0127] When the set of objects to be checked is empty, or all checked holes have been processed and passed the final deviation verification, the host computer uses the updated inspection data as the authoritative data source to directly generate a drilling / back-drilling machining program containing SpOF compensation information. This machining program includes the machining coordinates (original machining coordinates, SpOF compensation information, etc.) for each back-drilled hole. The generation process is directly linked to the inspection results after the check, without re-executing the complete process such as visual table processing, only performing necessary data format conversions and machining parameter mappings, thereby significantly shortening the closed-loop cycle from check to machining and ensuring that the manually corrected results are truly incorporated into the machining process.

[0128] S240: Issue target data and perform processing.

[0129] The host computer sends the generated machining program to the circuit board processing equipment. The processing equipment adjusts the drilling position according to the compensation amount in the program and processes the back drilled holes. In multi-axis scenarios, for example, the detection data of Z1 axis corresponds to Z#1, and the detection data of Z2 axis corresponds to Z#2. After the detection data of each axis is checked according to the correspondence between Z#n / #Zn, the target data is generated and sent out in a unified manner to ensure that the compensation data of each axis is correctly applied to the corresponding machining axis.

[0130] S260: Closed-loop connection of automated production lines.

[0131] When this application is applied to an automated production line, the host computer automatically locks the data path and processing task for the current batch of tests after inspection is completed. If the set of objects to be reviewed is not empty, the automatic transfer process automatically brings up the manual review interface; after the operator completes all reviews, the host computer automatically triggers the target data update and sends the updated target data to the processing equipment for CNC machining. This automated mode realizes an automated closed loop from inspection and review to processing, reducing manual intervention and making it suitable for high-volume, high-cycle production line environments.

[0132] Through the above-described processes S100 to S260, this application automatically corrects neighbor-correctable holes after hole scanning and inspection, and manually verifies or corrects holes without neighbor references, holes that fail to be identified, and holes whose concentricity deviation does not meet the standard. The application then generates back-drilling or compensation drilling files based on the verified authoritative data, thereby improving the concentricity of secondary drilling with the target hole center, shortening the closed-loop cycle, and facilitating quality traceability and process optimization.

[0133] The back-drill concentricity deviation compensation method and system provided in this application achieve the following technical effects through a closed-loop processing combining automatic compensation for through-hole position deviation with manual verification: By combining automatic neighborhood compensation with manual verification, the accuracy and reliability of concentricity deviation compensation are significantly improved; holes without neighborhood references, failed identification holes, and holes with unacceptable concentricity deviations are uniformly included in the verification object set for manual confirmation or correction, effectively eliminating blind spots in automatic correction and ensuring the integrity of processing data; the detection table is associated with the hole-level image, and the pixel deviation is converted into physical offset compensation based on calibration parameters and written back to the detection data, providing accurate quantitative basis for manual correction and improving the traceability of correction results; the target data for back-drill processing is directly generated using the verified detection data as the authoritative data source, eliminating the need to re-execute the complete visual table processing, significantly shortening the closed-loop cycle from verification to processing, and ensuring that the manual correction results can be effectively converted into processing parameters; it supports multi-axis detection data management, separate verification, and unified generation of target data, making it suitable for multi-axis processing scenarios and improving data management efficiency. Finally, theoretical coordinate data of the back drill hole is generated based on the compensated through hole position, thereby effectively improving the concentricity of the back drill hole and the through hole.

[0134] The back-drilling concentricity deviation compensation method provided in the embodiments of this application has been described in detail above, and specific examples have been used to illustrate the principle and implementation of this application. The description of the above embodiments is only for the purpose of helping to understand the technical solution and core idea of ​​this application. Those skilled in the art should understand that they can still modify the technical solutions described in the foregoing embodiments, or make equivalent substitutions for some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of this application.

Claims

1. A method for compensating for concentricity deviation in back drill biting, characterized in that, include: Acquire test data of the circuit board, the test data including the location information of the vias; Based on the detection data, the through holes are screened and their accuracy is determined to generate a set of verification objects; Determine whether the set of objects to be reviewed is empty; When the set of objects to be reviewed is empty, target data for back-drilling is generated based on the detection data, and the target data is sent to the circuit board processing equipment for processing. When the set of objects to be verified is not empty, the set of objects to be verified and its corresponding detection data are pushed to the human-machine interface, the through holes to be verified and their corresponding hole-level detection images are displayed, and manual operation is received until all the through holes in the set of objects to be verified are processed and all pass the final deviation verification. Then, the target data for back drilling is generated and sent to the circuit board processing equipment for processing.

2. The back-drill concentricity deviation compensation method according to claim 1, characterized in that, The process of generating the verification object set includes: dividing the through holes into a first type of hole and a second type of hole; and adding the second type of hole to the verification object set.

3. The back-drill concentricity deviation compensation method according to claim 2, characterized in that, The first type of hole includes: the through hole that is successfully identified, has a neighborhood reference condition, and has a concentricity deviation within a preset accuracy threshold range; the second type of hole includes: the hole that fails to be identified, the hole without a neighborhood reference, and the hole whose concentricity deviation does not meet the standard.

4. The back-drill concentricity deviation compensation method according to claim 3, characterized in that, The through holes that are not identified during the detection process are marked as the identification failure holes and added to the verification object set.

5. The back drill concentricity deviation compensation method according to claim 4, characterized in that, For the identified vias, a neighborhood range is defined, and neighboring vias of each via within the neighborhood range are retrieved. When no neighboring vias are retrieved or the number of neighboring vias is less than a preset value, the via is determined to be a via without a neighborhood reference and is added to the verification object set. When the number of neighboring vias retrieved is greater than or equal to the preset value, automatic neighborhood compensation is performed on the via.

6. The back-drill concentricity deviation compensation method according to claim 5, characterized in that, The automatic neighborhood compensation includes: statistically analyzing the concentricity deviation values ​​of all adjacent vias within the neighborhood range to determine a neighborhood concentricity deviation reference value; calculating the absolute difference between the components of the current via's concentricity deviation value and the neighborhood concentricity deviation reference value in each preset direction; when the absolute difference in any preset direction is greater than a preset difference threshold, correcting the concentricity deviation component of the current via in the preset direction to the corresponding component of the neighborhood concentricity deviation reference value; and when the absolute difference in all preset directions is not greater than the preset difference threshold, maintaining the original concentricity deviation value of the current via unchanged.

7. The back-drill concentricity deviation compensation method according to claim 5, characterized in that, After completing the automatic neighborhood compensation, the through holes are automatically judged for accuracy, including: calculating the concentricity deviation between the actual position and the theoretical position of each through hole; marking through holes with a concentricity deviation greater than a preset accuracy threshold as holes with unacceptable concentricity deviation and adding them to the verification object set; and classifying through holes with a concentricity deviation not greater than the preset accuracy threshold as the first type of hole.

8. The back-drill concentricity deviation compensation method according to any one of claims 1 to 7, characterized in that, After pushing and displaying the set of objects to be reviewed on the human-computer interaction interface, the method further includes: loading the set of objects to be reviewed and the corresponding detection data, and displaying the through holes to be reviewed in a list in sequence; in response to the selection of a through hole to be reviewed in the list, loading the hole-level detection image corresponding to the through hole, and superimposing a standard reference hole position and the current hole position to be corrected on the hole-level detection image; comparing the standard reference hole position and the current hole position to be corrected, and determining the type of manual review operation for the current hole position to be corrected, wherein the operation type includes confirmation, skipping, or correction.

9. The back-drill concentricity deviation compensation method according to claim 8, characterized in that, When the manual review operation type is correction, the step of determining the concentricity deviation compensation value includes: in response to the manual interactive adjustment operation of the current hole position to be corrected on the hole-level detection image, obtaining the pixel deviation component between the current hole position to be corrected and the standard reference hole position in a preset direction after adjustment; based on pre-calibrated parameters, converting the pixel deviation component into a physical offset compensation amount, and using the physical offset compensation amount as the concentricity deviation compensation value of the through hole.

10. The back-drill concentricity deviation compensation method according to claim 9, characterized in that, After determining the concentricity deviation compensation value, the method further includes: updating the physical offset compensation amount in the detection data of the via to update its concentricity deviation value, and marking the via as processed in the review object set; scanning the remaining states of the review object set to determine whether there are any unprocessed vias in the review object set; when there are unprocessed vias, returning to the step of sequentially displaying the vias to be reviewed in a list format; when there are no unprocessed vias, executing the step of sending the target data.

11. The back-drill concentricity deviation compensation method according to claim 9, characterized in that, The manual interaction adjustment operation includes: responding to the drag-and-move or point fine-tuning operation performed on the hole-level detection image to determine the final pixel position of the current hole to be corrected; when a skip instruction for the current through hole is received in the human-computer interaction interface, the current concentricity deviation value of the through hole is kept unchanged, and the through hole is directly marked as processed.

12. The back-drill concentricity deviation compensation method according to claim 10, characterized in that, Before sending the target data, a final deviation check is performed on all through holes marked as processed in the review object set: when a through hole marked as processed is checked and its current concentricity deviation value is still greater than the preset accuracy threshold, the marking status of the through hole in the review object set is reset to unprocessed, and the review information corresponding to the through hole is pushed back to the human-machine interface, and the manual correction step is re-executed; until all through holes in the review object set are marked as processed and the final check passes, the target data for back drilling is generated and sent.

13. A back-drill concentricity deviation compensation system, characterized in that, include: Circuit board processing equipment used to perform through-hole and back-drilling operations; A visual inspection device is used to inspect the through hole and output inspection data; The host computer is electrically connected to the vision inspection device; the host computer is electrically connected to the circuit board processing equipment; the host computer acquires the inspection data, generates target data for back drilling processing, and controls the circuit board processing equipment to process the back drill hole using the target data.

14. The back-drill concentricity deviation compensation system according to claim 13, characterized in that, The host computer is used to execute the back-drill concentricity deviation compensation method as described in any one of claims 1 to 12.

Images